Evidence-first automated news — publishing only what the evidence proves.
Coming SoonA fully automated AI news engine that collects events from multiple sources, verifies every claim against attributable evidence, and publishes articles written entirely from verified facts.
No article is published unless every core claim has primary-document backing or at least two independent confirmations.
Relevance is computed by a fixed, public formula — never by hidden editorial judgment. Every score is auditable.
Every published article links to original documents and shows exactly why it passed publication standards.
From raw signals to published article — a fully automated pipeline where every step is traceable.
Events are gathered from RSS feeds, news APIs, and open-source intelligence. Duplicates are automatically filtered.
Each event is broken into atomic factual claims. Every claim is verified against primary documents and independent sources.
Stories are scored using a transparent formula weighing evidence strength, impact, novelty, source diversity, and coverage volume.
Only stories meeting strict thresholds are published — with full evidence trails and source attribution.
The rules we never break.
We prefer delaying publication to publishing anything false, incomplete, or weakly sourced.
Decisions flow from disclosed machine rules, not subjective editorial calls or opinion.
The scoring formula and publication thresholds are public and versioned.
Every claim traces back to its original source documents for independent verification.